sequence复制ogg-01444 ora-20783错误

newway

错误内容：
2013-12-04 15:34:47  ERROR   OGG-01444  Error in replicating sequence value [ORA-20783:
ORA-08004: sequence SEQ_D_CERT_CERT_NO.NEXTVAL exceeds MAXVALUE and cannot be instantiated
ORA-06512: at "GOLDENGATE.REPLICATESEQUENCE", line 368
ORA-06512: at line 1, SQL BEGIN  "GOLDENGATE" .replicateSequence (TO_NUMBER(110108723), TO_NUMBER(10), TO_NUMBER(1), 'xxxxx', TO_N
UMBER(20), 'SEQ_D_CERT_CERT_NO', 'GOLDENGATE', TO_NUMBER(1), TO_NUMBER (0), ''); END;].
相关问题Maclean的文章里也有提到：
http://feed.askmaclean.com/archi ... alue-ora-20783.html
我查了一下OGG v11.2.1.0.3 release notes里说这个问题在该版本里已经解决。
11.2.1.0.13 Patch Set Bug Fixes

Total: 16  Bug Fixes


BUG:17439155 - EXTRACT ABENDS WITH DB2READLOG() ERROR
BUG:17434145 - 11.2.1.0.7+ EXTRACT IS SLOW WRITING TRAIL FILE, TOO MANY FSYNC CALLS
BUG:17268537 - MSSQL:UPGRADE:EXTRACT ABENDS WITH NO ERROR
BUG:17437023 - SQL/MX COLUMNS DEFINED AS CHAR/VARCHAR UCS2 HAVE CORRUPT DATA
BUG:17363190 - REPLICAT ABENDS: OGG-00551 DATABASE OPERATION FAILED: BINDING STATEMENT COLUMN
BUG:17355478 - EXTRACT ABENDS WITH "OUT OF ORDER SCN" ERROR AS OF A FIX IN BUG 16694692
BUG:17317448 - EXTRACT CAPTURES NULL FOR AN AFTER(NOT BEFORE) IMAGE OF A KEY COLUMN IN PKUPDATE
BUG:17300028 - REPLICAT ABENDS WITH ORA-20783 WHILE REPLICATING SEQUENCE
==================================================
因为权限问题，我无法获取BUG:17300028的具体描述，想确认一下是否就是这个bug和上述错误是一致的。

Liu Maclean(刘相兵

Hdr: 17300028 11.2.0.3.0 REPLICAT 12.1.2.0.0 PRODID-5757 PORTID-226
Abstract: REPLICAT ABENDS WITH ORA-20783 WHILE REPLICATING SEQUENCE

*** DPRASANN 08/12/13 08:22 am ***
@ SR (Mandatory for Customer bugs):
@ ---------------------------------
@ .
@ PROBLEM DESCRIPTION:
@ ---------------------
@ OGG Version :- OGGCORE_MAIN_PLATFORMS_130807.0056
@ RDBMS Version :- RDBMS_11.2.0.3.0_LINUX.X64_RELEASE with bp17 patch
@ .
@ While running sequence test tklmu80t.sql in short regress after fixing the
@ test framework for sequences it is seen that the replicat abends with the
@ following error.
@ .
@ From the sql file this is the DDL related to the failed sequence.
@ .
@ -- increment by 1 cache past maxvalue nocycle
@ CREATE SEQUENCE ID15 START WITH 6 MINVALUE 1 MAXVALUE 10 CACHE 6 NOCYCLE
@         INCREMENT BY 1;
@ .
@ REP1.rpt
@ ========
@ Wildcard MAP resolved (entry tklm*.*):
@   map "TKLMU80"."ID15", TARGET "TKLMU80"."ID15";
@ Resolving target sequence TKLMU80.ID15.
@ .
@ .
@ .
@ 2013-08-12 08:15:23  WARNING OGG-01004  Aborted grouped transaction on
@ 'TKLMU80.ID15', Database error 20783 ().
@ .
@ 2013-08-12 08:15:23  WARNING OGG-01003  Repositioning to rba 27146 in seqno
@ 0.
@ .
@ 2013-08-12 08:15:23  WARNING OGG-01154  SQL error 20783 mapping TKLMU80.ID15
@ to TKLMU80.ID15.
@ .
@ 2013-08-12 08:15:23  WARNING OGG-01003  Repositioning to rba 27882 in seqno
@ 0.
@ .
@ Source Context :
@   SourceModule            : [er.errors]
@   SourceID                :
@ [/scratch/aime1/adestore/views/aime1_adc4150392/oggcore/OpenSys/src/app/er/err
@ ors.cpp]
@ .
@ .
@ .
@                           :
@ [/net/adc6140839/scratch/dprasann/view_storage/dprasann_ogg1/work/worklv/oggor
@ a1/replicat(main+0x4a) [0x5e135a]]
@                           : [/lib64/libc.so.6(__libc_start_main+0xf4)
@ [0x336cc1d994]]
@                           :
@ [/net/adc6140839/scratch/dprasann/view_storage/dprasann_ogg1/work/worklv/oggor
@ a1/replicat [0x4ca649]]
@ .
@ 2013-08-12 08:15:23  ERROR   OGG-01444  Error in replicating sequence value
@ [ORA-20783:
@ ORA-8004: sequence ID15.NEXTVAL exceeds MAXVALUE and cannot be instantiated
@ ORA-6512: at "TKGGADMIN.REPLICATESEQUENCE", line 426
@ ORA-6512: at line 1 SQL BEGIN tkggadmin .replicateSequence (TO_NUMBER(11),
@ TO_NUMBER(20), TO_NUMBER(1), 'TKLMU80', 'ID15', UPPER('tkggadmin'),
@ TO_NUMBER(1), TO_NUMBER (0), ''); END;].
@ .
@ CUSTOMER IN PRODUCTION OR DEVELOPMENT OR QA?:
@ ---------------------------------------------
@ .
@ DIAGNOSTIC ANALYSIS:
@ --------------------
@ .
@ ERROR-STACK OBTAINED (xml/param/ggserror/archive.log  or rpt or Core/DDL
@ files):
@ ------------------------------------------------------------------------------
@ --
@ .
@ WORKAROUND:
@ -----------
@ .
@ REPRODUCIBILITY (Always, Sporadic, any peculiarity etc.):
@ ---------------------------------------------------------
@ .
@ TESTCASE (Location on bugftp):
@ ------------------------------
@ .
@ TESTCASE STEP-BY-STEP INSTRUCTIONS:
@ -----------------------------------
@ Steps to reproduce :-
@ .
@ From OGGCORE_MAIN_PLATFORMS_130807.0056 label view or latest copy all the
@ test files from bugsftp to tklocal of this view and run the test as shown
@ below.
@ .
@ setenv GGNOCLEANUP true
@ setenv XOUT_EXTRACT true
@ setenv TKLM_TEST_FILE tklmu80
@ oratst tkmain_gglm2
@ .
@ RELATED BUGS (Bugid's if there are any known bugs related to this bug):
@ -----------------------------------------------------------------------
@ .
@ 24 HOUR CONTACT INFORMATION FOR P1 BUGS (Mandatory for P1 bugs only):
@ ---------------------------------------------------------------------
@ .
@ ADDITIONAL INFORMATION/KEYWORDS (Any additional info not covered above):
@ ------------------------------------------------------------------------
@ .
@ *** DPRASANN 08/12/13 08:23 am *** (CHG: FixBy->12.1.2.0.0)
*** DPRASANN 08/12/13 08:23 am *** (CHG:3rd Party/QA->DPRASANN)
*** DPRASANN 08/12/13 08:23 am *** (CHG: Asg->ABROWN)
*** DPRASANN 08/12/13 08:25 am ***
@ Attachment:tkggadt1_00.prm:File Uploaded through Bugsftp
@ Attachment:tkggadt1_00cdb.prm:File Uploaded through Bugsftp
@ Attachment:tkggadt1_00ir.prm:File Uploaded through Bugsftp
@ Attachment:tkggconfigprm.tsc:File Uploaded through Bugsftp
@ Attachment:tkgglm2.tsc:File Uploaded through Bugsftp
@ Attachment:tkgglm2_sprun.tsc:File Uploaded through Bugsftp
*** DPRASANN 08/12/13 08:26 am ***
@ Attachment:tkggutil.sql:File Uploaded through Bugsftp
*** DPRASANN 08/12/13 08:26 am ***
@ Attachment:tkggfetchcols.log:File Uploaded through Bugsftp
*** DPRASANN 08/12/13 08:30 am ***
@ After this bug is fixed please enable the test by uncommenting the the call
@ to tklmu80 in test/tkgg/src/tkgglm_run2.tsc as shown below
@ .
@ tkgglm_run2.tsc
@ ===============
@    #;-- - SEquence test
@    #;-- BUG 17300028
@    #runtest tkgglm2 tklm_test=tklmu80
@ .
@ to
@ .
@ tkgglm_run2.tsc
@ ===============
@    #;-- - SEquence test
@    runtest tkgglm2 tklm_test=tklmu80
*** DPRASANN 08/12/13 08:34 am ***
@ Attachment:oggora1.tar.gz:File Uploaded through Bugsftp
*** ABROWN 08/12/13 11:50 am *** (CHG: Asg->SMIJATOV)
*** ABROWN 08/12/13 11:50 am ***
@ Serge: As per discussion in today's meeting.
*** SMIJATOV 08/14/13 10:39 pm ***
@ The issue here is that replicat sequence code tries to move "ahead" of the
@ source, always. When the end of the possible sequence space has reached,
@ there is no "ahead", so we need to handle that.
@ .
@ I am working on it now.
@ *** BGARIN 08/15/13 11:03 am *** (CHG: FixBy->12.1.2.1.0)
*** BGARIN 08/15/13 11:03 am ***
@ Serge confirmed that the same behavior exists in 11.2.1.0.6 so I am deferring
@ this to 12.1.2.1.0 since this is an edge case and is NOT a regression.
*** AUTOREP 08/19/13 09:56 am ***
@ The URL for the code review for this transaction is
@ https://orareview.us.oracle.com/transaction/smijatov_bug-17300028
*** SMIJATOV 08/21/13 09:57 am ***
@ ADE Transaction Id: smijatov_bug-17300028
@ FILES:
  
@ oggcore/OpenSys/redist/SQLScripts/sequence.sql@@/main/7

*** SMIJATOV 08/21/13 09:59 am ***
RELEASE NOTES:
  
]]Sequence replication fails in replicat when sequence reaches MAXVALUE for non
]]-cyclic sequences.
  
@ INTERNAL PROBLEM DESCRIPTION:
  
@ An attempt is made to advance sequence beyond its maximum value.
  
@ INTERNAL FIX DESCRIPTION:
  
@ Handle an error resulting from sequence advanced beyond its maximum value.
  
@ BACKPORT FEASIBLE:
  
@ Yes
  
@ FORWARD MERGE REQUIRED:
  
@ No (merged to main branch)
  
REDISCOVERY INFORMATION:
  
Test case in the bug.
  
WORKAROUND:
  
None
  
*** ADEBUG 08/21/13 09:59 am *** (CHG: Fixed->12.1.1.0.0
)
*** ADEBUG 08/21/13 09:59 am *** (CHG: Sta->80)
*** ADEBUG 08/23/13 04:01 am ***
@ OGGCORE_MAIN_PLATFORMS_130822.1405 contains changes from smijatov_bug-1730002
@ 8
*** ARU 09/12/13 05:12 am ***
@  *** (CHG: Cumulative Patch Inclusion -> NULL->OGGBP 11.2.1.0.0 Requested)
@ To view the CPM request created, refer to
@http://aru.us.oracle.com/ARU/CIView/process_form?rids=34857
@ Series name: OGG Bundle Patch 11.2.1.0.0
*** ARU 09/12/13 05:18 am ***
@ CUMULATIVE PATCH INCLUSION REQUEST #2233549 CREATED IN BUG 17448620 FOR
@ 11.2.1.0.13OGGBP
*** ARU 09/12/13 05:18 am ***
@  *** (DEL: Cumulative Patch Inclusion -> OGGBP 11.2.1.0.0 Requested)
@  *** (CHG: Cumulative Patch Inclusion -> NULL->OGGBP 11.2.1.0.13 Approved, In
@ Process)
*** SMIJATOV 09/17/13 11:39 pm ***
@ ADE Transaction Id: smijatov_ci_backport_17300028_11.2.1.0.13oggbp
@ [CI backport ]
@ FILES:
  
@ oggcore/OpenSys/redist/SQLScripts/sequence.sql@@/st_oggcore_11.2.1/1

*** SMIJATOV 09/17/13 11:39 pm ***
@ ADE Transaction Id: smijatov_ci_backport_17300028_11.2.1.0.13oggbp
@ [CI backport ]
@ FILES:
  
@ oggcore/OpenSys/redist/SQLScripts/sequence.sql@@/st_oggcore_11.2.1/1

*** DPRASANN 09/25/13 11:07 pm *** (CHG: Fixed->12.1.1.0.0)
*** DPRASANN 09/25/13 11:07 pm *** (CHG: Sta->90)
*** ARU 10/15/13 02:02 am ***
@  *** (DEL: Cumulative Patch Inclusion -> OGGBP 11.2.1.0.13 Approved, In
@ Process)
*** ARU 10/15/13 02:02 am ***
@  *** (CHG: Cumulative Patch Inclusion -> NULL->OGGBP 11.2.1.0.13 Released)

newway

OK：The issue here is that replicat sequence code tries to move "ahead" of the
source, always. When the end of the possible sequence space has reached, there is no "ahead", so we need to handle that.目前oggv11.2.1.0.13之后的发布的版本还主要是11g的，唯一发布了针对10g的还下载不了，受限制，只能先等等了，人为干预这种非循环sequence,到最大值的情况了

lunar

sequence的复制可以workround吧，复制的时候不要sequence，然后写脚本后台近期更新（适具体情况确定job执行的频繁程度），因为sequence的复制貌似再切换时才有用，平时没啥用吧。。。我的客户就是这样的，为此做了切换测试